Deepfakes: What are they and why they matter

IMAGE CREDIT:
Image credit
iStock

Deepfakes: What are they and why they matter

Deepfakes: What are they and why they matter

Subheading text
Deepfakes can be used to slander and misrepresent individuals and corporations. But with proper knowledge, executives can protect themselves and their businesses.
    • Author:
    • Author name
      Quantumrun Foresight
    • January 19, 2022

    Insight summary



    Deepfakes, a technology that allows the creation of hyper-realistic fake videos or audio, emerged in 2017 and has since sparked both concern and opportunity. While the technology has been misused to create deceptive content, it also offers potential benefits such as enhancing online privacy, transforming advertising strategies, and aiding in law enforcement. However, the long-term implications of deepfakes are vast, ranging from the need for digital literacy education and the growth of a new industry focused on deepfake detection, to ethical considerations and increased energy consumption.



    Deepfakes context



    The term "deepfake" entered public awareness in 2017 when a Reddit user shared pornographic clips that used open-source face-swapping technology. In these videos, they swapped the faces of celebrities like Scarlett Johansson, Taylor Swift, Gal Gadot, and others with pornographic performers. This was just the beginning.



    Deepfake technology allows people to create video or audio of events that never occurred. For example, celebrities and politicians have found themselves in videos doing and saying things they never really did or said. Concerns about fake visuals and audio created by deepfake technologies led to a proliferation of countermeasures. To eliminate the negative impacts of deepfakes, it is essential to introduce new laws. In 2020, social media platforms including Twitter and Facebook banned deepfakes from their networks. 



    It takes a few steps to create a high-quality deepfake clip. First, run thousands of face shots of two persons through an encoder. The encoder highlights similarities between the two faces and lowers them to shared features by compressing the images. Then, the faces are recovered from compressed photos using a decoder. As faces are different, one decoder is trained to recover the first person's face and another to recover the second person's face. After that, the creator needs to feed encoded pictures into the "wrong" decoder to apply the face swap. 



    Disruptive impact



    Deepfakes, while posing significant threats, also present unique opportunities. For individuals, the ability to create realistic digital personas could transform online interactions. For instance, a person could use a deepfake to maintain privacy during video calls, presenting a digital avatar instead of their real face. This feature could be particularly useful in situations where personal safety or anonymity is paramount.



    For corporations, deepfakes could redefine advertising and customer engagement strategies. Companies could create virtual spokespersons, tailored to resonate with different target audiences. This strategy could lead to more personalized and effective marketing campaigns. However, this also raises ethical considerations, as consumers may be deceived by hyper-realistic yet artificial representations.



    Governments could use deepfake technology for public safety and security purposes. For example, law enforcement agencies could use deepfakes in sting operations, creating realistic scenarios without putting human officers at risk. However, the potential for misuse is high, and it is crucial for governments to establish strict regulations to prevent the abuse of this technology. In the long run, the impact of deepfakes will largely depend on how responsibly we use and regulate this powerful tool.



    Implications of deepfakes



    Wider implications of deepfake technology may include: 




    • Its use in law enforcement to reconstruct crime scenes with the interrelationship of temporal and spatial artifacts. 

    • Its use by fashion retail businesses to create virtual trial rooms for customers to try favorite products without actually trying them.

    • Providing individuals with new tools for integration and self-expression in the online world. For instance, individuals could create avatars of themselves for online self-expression.

    • The advanced and pervasive use of deepfakes across the media by multiple bad actors. In this worst-case scenario, deepfakes may impair society’s ability to believe in what they see and hear, making broader swathes of society more susceptible to propaganda and various forms of manipulation.

    • A surge in the demand for digital literacy education, resulting in a more informed and discerning populace that can better distinguish between real and manipulated content.

    • A new industry focused on deepfake detection and prevention, creating job opportunities and economic growth.

    • Advancements in artificial intelligence and machine learning, pushing the boundaries of what is technologically possible.

    • An increase in energy consumption, as the creation and detection of deepfakes require significant computational resource.



    Questions to consider




    • What are the potential impacts of deepfakes on society?

    • Do you think laws introduced by governments will help eliminate the negative applications of deepfakes? 

    • What innovations can deepfake technology be applied to in the future?


    Insight references

    The following popular and institutional links were referenced for this insight: